Two mitochondrial (CytB and COII) and one nuclear gene (Gpi) were selected (i) among those most widely used in the field, (ii) of single copy for the nuclear one, and (iii) presenting common sequences of sufficient size for applying phylogenetic tools. They were analyzed using maximum likelihood trees and phylogenetic networks. Remarkably, only three significant clusters instead of seven were found with the mitochondrial genes. With the nuclear gene, surprisingly, all seven expected clusters did not have significant bootstrap values. Moreover, DTUs TcV and TcVI were indistinguishable as were TcIII and TcIV. Additionally, we have undertaken a minireview of seventy-five publications presenting phylogenetic trees with identifiable DTUs that allowed us, together with our own results, to seriously question the structuring of T. cruzi into six or seven separated DTUs.
We propose that mitochondrial typing in three clusters currently named mtTcI, mtTcII, and mtTcIII is robust whereas nuclear typing may lead to a questionable clustering but it is valuable for detecting mitochondrial introgression, heterozygous states and allelic composition.